Upload folder using huggingface_hub

app.py

@@ -1,13 +1,9 @@
 import gradio as gr
-from apps import pdf_cropper, text_tools
-
-# --- 主程序入口 ---
-# 这里是“应用集市”的容器。
-# 每次添加新工具，只需要 import 进来，并在 tab_list 中注册即可。
+from apps import pdf_cropper, text_tools, text_diff
 
 def create_main_interface():
     with gr.Blocks(title="我的科研工具箱") as main_app:
-        gr.Markdown("# 🛠️ 科研效率工具箱")
+        gr.Markdown("# 🛠️ 写作妙妙小工具")
         
         # 使用 Tab 布局来切换不同的工具
         with gr.Tabs():
@@ -19,6 +15,10 @@ def create_main_interface():
             # --- 工具 2: 文本分析 (示例) ---
             with gr.TabItem("📝 文本统计"):
                 text_tools.create_ui()
+
+            # --- 工具 3: 文本比对 ---
+            with gr.TabItem("🔍 文本比对"):
+                text_diff.create_ui()
                 
             # --- 可以在这里继续添加更多 Tab ---
             
@@ -26,4 +26,14 @@ def create_main_interface():
 
 if __name__ == "__main__":
     app = create_main_interface()
-    app.launch(inbrowser=True)
+    
+    # 创建一个微调过的 Soft 主题
+    # primary_hue="indigo" (靛青色，更有科技感)
+    # neutral_hue="slate" (岩灰色，更护眼)
+    custom_theme = gr.themes.Soft(
+        primary_hue="indigo",
+        neutral_hue="slate",
+    )
+    
+    # 注意：在 Gradio 新版本中，theme 参数已移动到 launch() 方法中
+    app.launch(inbrowser=True, theme=custom_theme)

apps/pdf_cropper.py

@@ -3,6 +3,8 @@ from PIL import Image, ImageChops
 import img2pdf
 import io
 
+import os
+
 # --- 核心逻辑 (复用之前的代码) ---
 def trim_whitespace(im, fuzz_level=20):
     bg = Image.new(im.mode, im.size, (255, 255, 255))
@@ -15,10 +17,12 @@ def trim_whitespace(im, fuzz_level=20):
 
 def process_pipeline(files, quality, fuzz_level, progress=gr.Progress()):
     if not files: return None
-    pdf_components = []
+    output_files = []
+    
     for filepath in progress.tqdm(files, desc="处理中"):
         try:
             with Image.open(filepath) as img:
+                # 处理透明背景
                 if img.mode in ('RGBA', 'LA') or (img.mode == 'P' and 'transparency' in img.info):
                     bg = Image.new('RGB', img.size, (255, 255, 255))
                     if img.mode!= 'RGBA': img = img.convert('RGBA')
@@ -27,17 +31,26 @@ def process_pipeline(files, quality, fuzz_level, progress=gr.Progress()):
                 else:
                     img = img.convert('RGB')
                 
+                # 裁边
                 trimmed = trim_whitespace(img, fuzz_level)
-                byte_arr = io.BytesIO()
-                trimmed.save(byte_arr, format='JPEG', quality=int(quality))
-                pdf_components.append(byte_arr.getvalue())
-        except Exception: continue
+                
+                # 获取原始文件名并构建输出路径
+                original_name = os.path.basename(filepath)
+                name, _ = os.path.splitext(original_name)
+                # 保持文件名，改为 .pdf 后缀
+                output_filename = f"{name}_cropped.pdf"
+                
+                # 保存为 PDF
+                # img2pdf 需要 bytes 或文件路径，这里我们先存为临时图片再转，或者直接用 PIL save pdf
+                # PIL save PDF 支持单张
+                trimmed.save(output_filename, "PDF", resolution=100.0, save_all=True)
+                output_files.append(output_filename)
+                
+        except Exception as e:
+            print(f"Error processing {filepath}: {e}")
+            continue
 
-    if not pdf_components: return None
-    output_filename = "output_document.pdf"
-    with open(output_filename, "wb") as f:
-        f.write(img2pdf.convert(pdf_components))
-    return output_filename
+    return output_files
 
 # --- 模块化 UI 接口 ---
 def create_ui():
@@ -48,11 +61,11 @@ def create_ui():
     """
     with gr.Row():
         with gr.Column():
-            file_input = gr.File(file_count="multiple", file_types=["image"], label="上传图片")
-            quality = gr.Slider(10, 100, 90, label="质量")
+            file_input = gr.File(file_count="multiple", file_types=["image"], label="上传图片 (支持多选)")
+            quality = gr.Slider(10, 100, 90, label="质量 (仅用于压缩，当前直接转PDF可忽略)")
             fuzz = gr.Slider(0, 100, 30, label="容差")
             btn = gr.Button("开始处理", variant="primary")
         with gr.Column():
-            output = gr.File(label="下载 PDF")
+            output = gr.File(label="下载结果", file_count="multiple")
             
     btn.click(process_pipeline, [file_input, quality, fuzz], output)

apps/text_diff.py

@@ -0,0 +1,52 @@
+import gradio as gr
+from difflib import Differ
+
+def diff_texts(text1, text2):
+    d = Differ()
+    # 使用 splitlines(keepends=True) 可以保留换行符，使 diff 更自然
+    # 但 HighlightedText 通常基于 token/word，这里我们按单词分割来做细粒度对比
+    # 如果需要按行对比，可以调整逻辑
+    
+    # 这里演示按单词/字符流的对比，效果类似 Code Diff
+    return [
+        (token[2:], token[0] if token[0] != " " else None)
+        for token in d.compare(text1.splitlines(keepends=True), text2.splitlines(keepends=True))
+    ]
+
+def diff_texts_word_level(text1, text2):
+    d = Differ()
+    # 按单词分割，更适合文章类文本
+    diff = d.compare(text1.split(), text2.split())
+    result = []
+    for token in diff:
+        code = token[0]
+        word = token[2:] + " " # 补回空格
+        if code == " ":
+            result.append((word, None))
+        elif code == "-":
+            result.append((word, "-")) # 删除
+        elif code == "+":
+            result.append((word, "+")) # 新增
+    return result
+
+def create_ui():
+    with gr.Row():
+        with gr.Column():
+            t1 = gr.Textbox(label="原始文本 (Original)", lines=15, placeholder="输入旧版本文本...")
+        with gr.Column():
+            t2 = gr.Textbox(label="新文本 (New)", lines=15, placeholder="输入新版本文本...")
+    
+    with gr.Row():
+        btn = gr.Button("🔍 对比差异", variant="primary")
+    
+    # HighlightedText 组件非常适合展示 Diff
+    # color_map 定义了不同标记的颜色：+ 为绿色(新增)，- 为红色(删除)
+    diff_output = gr.HighlightedText(
+        label="差异视图 (Diff View)",
+        combine_adjacent=True,
+        show_legend=True,
+        color_map={"+": "green", "-": "red"}
+    )
+    
+    # 这里使用 word level diff，视觉效果更好
+    btn.click(diff_texts_word_level, inputs=[t1, t2], outputs=[diff_output])